The debate is running again on LinkedIn. This time, the argument is between AI First and Data First. The Data First camp argues that AI initiatives built on poor data foundations will collapse. You have to get your data house in order before AI can deliver value. The AI First camp argues that leading with AI capability is the key to staying competitive. Both sides are making real points.

Both sides are also missing the more important one.

Neither AI First nor Data First survives without a customer to serve. The debate over which comes first between two enabling technologies is entirely the wrong argument. It is a conversation about the engine before anyone has agreed on where the car is going.

Customer First. Data second. AI last.

That is the sequence that produces outcomes rather than initiatives.

Why the Buzzterm Debate Keeps Happening

Mobile First changed how design teams thought about interfaces. Digital First shifted organizational investment toward digital channels. Cloud First redirected infrastructure spending. Each of these “First” labels served a real purpose at a specific moment. They created focus. They gave teams a shared mental model for prioritizing decisions in a period of technological transition.

The problem with “First” thinking is that it is inherently means-driven. It describes how you build. It omits why you build. And when the “First” label gets attached to a technology rather than a purpose, the technology becomes the destination rather than the vehicle.

AI First says: design your products and workflows so that AI is native to them, not an afterthought. This is a reasonable design principle applied to the components where AI genuinely belongs.

Data First says: build a reliable, well-governed, high-quality data infrastructure before investing in AI. This is a reasonable engineering discipline applied to any initiative where data quality determines outcome quality.

Both are correct as principles. Neither should be the governing philosophy of an organization. Only one thing should occupy that position, and it has nothing to do with technology.

The Sequence That Actually Works

Data empowers your decisions to serve your customers. AI leverages data to create the experiences your customers expect. Both data and AI serve the customer. Without customers, your data and AI initiatives are pointless.

Read that sequence again. Data is in the service of decisions about customers. AI serves customer experiences. The customer is the reason both exist. Remove the customer from the equation, and you have a sophisticated infrastructure serving no particular purpose.

This is the sequence that produces real outcomes.

Start with the customer problem. Who is the person with this problem? What does solving it actually look like from their perspective? What does success feel like for them? These are the questions that define what you are building toward.

Then ask what data you need to understand that problem deeply enough to solve it reliably. A robust data infrastructure can improve the quality of decisions about customers. A data initiative with no traceable connection to a customer decision is infrastructure for its own sake.

Then ask whether AI is the right tool for this specific problem. Sometimes it is. Often, it is the slower, more expensive, less reliable path to the customer outcome. I consulted with a company last week that could have achieved the same outcomes faster and cheaper using a simple search tool rather than implementing AI. The AI First mindset made AI feel obligatory. The customer did not need it.

Customer First. Data second. AI, maybe.

Another Consideration of Priorities

I’d argue, though, that before data, there are other factors that are more important.

People. Culture. Empathy. Aspirations.

The employees who understand the customer problem deeply enough to define it accurately. The culture that treats customer empathy as a design requirement rather than a value statement. The shared understanding of what a good experience actually feels like, developed through genuine human observation, well before any technology decision is made.

These are the assets that make data meaningful and AI useful. An organization that jumps from customer problem to data strategy has skipped the human infrastructure that makes both work. Employees who know the customer, designers who can translate that knowledge into clear problem definitions, and a culture that holds the customer’s reality as the primary constraint, these belong in the sequence between the customer problem and the data that serves it.

The full sequence, honestly stated, is: Customer. Employees and Culture. Design and Empathy. Data. Technology last. AI, maybe, as one technology among several.

Every “First” label that attaches to a technology belongs somewhere in that last position.

The Cultural Consequence of Getting This Wrong

Where a company places “First” has consequences that run deeper than design principles.

An AI First company hires for AI capability and measures AI performance. Model accuracy. Inference throughput. Token efficiency. These are real metrics. They tell you how the machine is performing. They tell you almost nothing about whether customers are better served.

A Customer First company measures customer outcomes and traces enabling factors back from there. Customer retention. Customer effort. Net Promoter Score. AI performance sits within those outcomes as one contributing variable among several. The technology is evaluated by its contribution to the customer result rather than in isolation.

The roadmap difference is even more consequential. An AI First roadmap builds features that demonstrate AI capability and trusts that customer value will follow. A Customer First roadmap starts with customer outcomes and works backward to the enabling features, which may or may involve AI. The difference in what gets built and what gets cut is substantial.

A company that has adopted AI First as its governing philosophy has placed technology in the customer’s position. Every design decision, every resource allocation, every prioritization question gets filtered through what is best for the AI initiative. The customer’s reality becomes a secondary input into a process that began elsewhere.

This is observable. It is visible in products that are impressively capable and persistently frustrating to use. In features that work elegantly for the AI’s processing model and awkwardly for the human’s cognitive model. Roadmaps that celebrate model improvements while customer satisfaction trends in a different direction.

The AI Ideology Problem

In Infailible, I describe AI as an ideology. The belief in AI has grown to the point where it shapes decisions independent of evidence. Organizations deploy AI because they believe in it, because their competitors are deploying it, because investors expect it, and because their brand narrative requires it. The deployment decision comes before the problem definition. The technology comes before the customer.

This is ideology in the precise sense. A belief system that organizes behavior, independent of whether the beliefs are producing the outcomes they claim to produce.

The Data First argument, well-intentioned as it is, fails to escape this dynamic. It remains a technology argument. It says you should prioritize data before AI, which is absolutely correct as a technical discipline, but it still positions the technology conversation as the primary one. The customer conversation is assumed to be taking place elsewhere.

It rarely happens somewhere else. For many organizations, it is happening at all barely. The debate about which technology comes first is a proxy for the harder question of whether the organization is genuinely organized around the problems of the people it serves.

One First

There can only be one “First” in an organization. It is the lens through which decisions get made. It is the tiebreaker when resources are contested, priorities conflict, and the roadmap has to be cut.

AI First and Data First are both useful frameworks for components of the work. AI-first thinking applied to the parts of a product where AI genuinely belongs produces better AI-native design. Data-First thinking applied to infrastructure decisions produces stronger data foundations.

Neither belongs in the governing position. That position belongs to the customer. Always. Regardless of how impressive the technology cycle is, and this one is genuinely impressive, the organization that loses the customer as its primary orientation loses the reason it exists.

Customer First means understanding the problem before designing the solution. It means evaluating AI and data by their measurable contribution to customer outcomes rather than by their technical sophistication. It means being willing to recommend a search tool when that’s what the customer problem actually requires.

Customer First. Data to understand and serve them. AI to create the experiences they expect, when AI is genuinely the right tool to do it.

That is the sequence. There is no tie.

---

If you find this content valuable, please share it with your network.

Follow me for daily insights.

Book me to speak at your next event.

Start managing your agents for free.

Chris Hood is an AI strategist and author of the #1 Amazon Best Seller Infailible and Customer Transformation, and has been recognized as one of the Top 30 Global Gurus for Customer Experience. His latest book, Unmapping Customer Journeys, is available now!